Children’s Productive Use of Academic Vocabulary · Keywords: productive use of academic...

Running head: CHILDREN’S PRODUCTIVE USE OF ACADEMIC VOCABULARY 1

To appear in Discourse Processes

Children’s Productive Use of Academic Vocabulary

Shufeng Ma1, Jie Zhang2, Richard C. Anderson1,

Joshua Morris1, Kim Thi Nguyen-Jahiel1, Brian Miller3, May Jadallah4, Jingjing Sun5,

Tzu-Jung Lin6, Theresa Scott7, Yu-Li Hsu1, Xin Zhang1, Beata Latawiec8, Kay Grabow9

1University of Illinois at Urbana-Champaign

2Western Kentucky University

3Towson University

4Illinois State University

5University of Montana

6The Ohio State University

7Washington Elementary School, Champaign, Illinois

8Wichita State University

9Thomas Paine Elementary School, Urbana, Illinois

Author note:

The research reported in this paper was supported by the Institute of Education Sciences, U.S. Department

of Education, through Grant R305A080347 to the University of Illinois at Urbana-Champaign, Richard C.

Anderson, Principal Investigator. The opinions expressed are those of the authors and do not represent

views of the Institute or the U.S. Department of Education. The authors are pleased to acknowledge the

contributions of Shengcheng Yuan, Aini Marina Ma’rof, Nikisha Blackmon, and Xiaoying Wu.

Corresponding authors:

Shufeng Ma, M.A.

Center for the Study of Reading

University of Illinois at Urbana-Champaign

157 Children’s Research Center

51 Gerty Drive

Champaign, IL 61820

217-898-2580 (voice)

[email protected]

Jie Zhang, Ph.D.

3088 Gary A. Ransdell Hall

Department of Educational Administration,

Leadership, and Research

Western Kentucky University

Bowling Green, KY 42101

270-745-2933 (voice)

[email protected]

mailto:[email protected]

mailto:[email protected]

CHILDREN’S PRODUCTIVE USE OF ACADEMIC VOCABULARY 2

Abstract

Instructional influences on productive use of academic vocabulary were investigated among 460 mostly

African American and Latina/o fifth graders from 36 classrooms in eight public schools serving low-

income families. Students received a six-week unit on wolf management involving collaborative group

work (CG) or direct instruction (DI). The big question that students tried to answer during the unit was

whether a community should be permitted to destroy a pack of wolves. In an individual oral interview

about an analogue to the wolf question, whether whaling should be allowed, both CG and DI students

used more general and domain-specific academic vocabulary from the Wolf Unit than uninstructed

control students. CG students used more general academic vocabulary in the whale interview than DI

students, and this was mediated by the CG students’ greater use of general academic vocabulary in

classroom dialogue during the Wolf Unit. These results suggest that collaborative group work is an

effective instructional approach to promote acquisition and productive use of academic vocabulary for

children from underserved communities.

Keywords: productive use of academic vocabulary, collaborative group work, direct instruction


Children’s Productive Use of Academic Vocabulary

Mastery of academic vocabulary is necessary for comprehension of content area textbooks,

technical material in trades and professions, and newspaper articles on current events, and is no doubt

important for thinking, speaking, and writing about all manner of topics. In research on growth in

knowledge of academic vocabulary, the test for whether or not students know academic words has almost

always involved recognizing the words or selecting definitions of the words from among several choices.

Almost no research has been done on productive use of academic vocabulary words (Pearson, Hiebert, &

Kamil, 2007).

In a productive use of a vocabulary item, an individual is able to say or write the word, in contrast

to merely being able to recognize it. As we will use the term, productive use has the further requirement

of a spontaneous or unprompted use. Thus, although a task in which children rapidly name common

objects meets one requirement of productive use of words, the pictures of the objects, and the instructions

to name them, strongly prompt the words. The meaning of the term productive use is satisfied when a

child [1] articulates or writes a word and [2] does so spontaneously, as when telling a story or writing an

essay. A test for productive use of a word need not be ‘natural’ in the sense, for instance, of overhearing a

child use the word on the playground. The use can be elicited as long as the child has plenty of degrees of

freedom in how to respond. Ideally, the elicitation would afford but not require use of the word. A robust

assessment of productive use will evaluate whether the child can use the word in a context different from

the one in which it was acquired. The term productive use is comparable to Laufer and Nation’s (1999)

term free productive use which they contrasted with controlled productive use. Supplying a suitable word

in the following context is an example of controlled productive use: “The garden was full of fra_____

flowers.”

The present study examined instructional influences on productive use of academic vocabulary.

The influence of two methods, direct instruction and collaborative group work, was compared with

African American and Latina/o children, the two largest groups of underserved children in the United

States. The National Assessment of Educational Progress reports that these students continue to lag in


reading and writing (Salahu‐Din, Persky, & Miller, 2008). Students from low income, minority homes

often lack familiarity with lexical, grammatical, and discourse features of an academic voice (Scarcella,

2003; Snow & Uccelli, 2009). Among underserved children who are second language learners, there is a

large gap between receptive language (listening and reading) and expressive language (speaking and

writing), with expressive language lagging as much as one standard deviation behind receptive language,

according to one recent study (Keller, Troesch, & Grob, 2015). Hence, there is a pressing need to identify

instructional methods for underserved children that expand knowledge of academic vocabulary words and

create functional contexts for productive use of the words.

The challenge of academic vocabulary

Academic vocabulary is an important component of academic language, a register of English

used in schools and universities that is critical for academic success (Corson, 1997; Scarcella, 2003;

Snow, 2010). Academic vocabulary can be classified into two categories: domain‐specific vocabulary,

also called technical vocabulary (Hiebert & Lubliner, 2008) or Tier 3 words (Beck, McKeown, & Kucan,

2002), and general academic vocabulary (e.g., Bailey, 2006; Hiebert & Lubliner, 2008). Domain-specific

vocabulary refers to technical words that are necessary for understanding and expressing key concepts

within a domain, such as mean and standard deviation within the domain of statistics. General academic

vocabulary refers to broadly useful words that appear in the discourse of many disciplines, such as affect,

decline, and provide.

Knowing domain‐specific vocabulary is indispensable for proficiency in science and other

technical subjects. Domain-specific words pose a challenge for students because these words tend to be

labels for unfamiliar concepts (Bravo & Cervetti, 2008). While domain-specific vocabulary is challenging

for all students, general academic vocabulary also presents a challenge for English language learners and

students from low-income and less-educated families. These students depend on schools for exposure to

academic words, which are “usually non‐concrete, low in imagery, low in frequency, and semantically

opaque” (Corson, 1997, p.696), and therefore difficult to learn.


Incidental learning of words while reading plays an essential role in vocabulary growth (Nagy &

Anderson, 1984; Stanovich, 1986). However, research suggests that level of word abstractness influences

vocabulary learning (Schwanenflugel et al., 1997), which makes acquisition of academic vocabulary more

demanding than acquisition of other words. There is little evidence of incidental learning of complex and

abstract words from one or two exposures during normal reading (Nagy, Anderson, & Herman, 1987).

Because academic vocabulary is often complex and abstract, studies suggest that incidental learning is not

enough for students to acquire new academic words among second language learners and other students

with limited exposure to academic discourse (Carlo et al., 2004). Many have argued that to accelerate

these students’ growth in literacy and content knowledge, academic vocabulary should be taught

explicitly (e.g., Snow, 2010).

In recent years, there has been an increase in research evaluating instructional interventions

designed to teach academic vocabulary to students from linguistically diverse backgrounds. Some

interventions have focused on explicit teaching of sets of generally useful academic vocabulary words

(e.g., Beck, McKeown, & Kucan, 2002; Carlo et al., 2004; Snow, Lawrence, & White, 2009; Lesaux,

Kieffer, Faller & Kelley, 2010; Townsend & Collins, 2009), while others have integrated vocabulary

teaching into science or literacy instruction (e.g., August, Branum‐Martin, Cardenas‐Hagan, & Francis,

2009; Vaughn et al., 2009). These interventions show some promising effects on English vocabulary

knowledge and reading comprehension (although see Marulis & Neuman, 2010), as well as on content

knowledge, regardless of the student’s first language.

Academic vocabulary acquisition through classroom discussion

Research suggests that vocabulary learning is most likely to happen when students have multiple

exposures to vocabulary words in varied and meaningful contexts (e.g., Kelley, Lesaux, Kieffer, & Faller,

2010). An interactive learning format, therefore, may increase students’ chances of encountering and

using academic vocabulary and provide more opportunities for them to negotiate meanings in different

contexts (Oxford, 1997). In a review of effective approaches to language instruction, Ellis (2005) pointed

out that, as compared to direct instruction, small group work increases classroom talk and involves


students in a greater variety of speech acts. Mol, Bus, and de Jong (2009) found that an interactive

storybook reading program resulted in an 8% increase in children’s expressive vocabulary. Carlisle,

Fleming, and Gudbrandsen (2000) reported that involvement in discussion and hands‐on activities in

fourth and eighth grade science led to vocabulary growth, particularly when students already had partial

knowledge of the words.

Classroom discussion may facilitate the acquisition of academic vocabulary because discussion

often creates a high level of involvement, perhaps leading to deep processing of information. Discussion

requires understanding words in context, supplementing the understanding that can be gleaned from

definitions. Barron and Melnick (1973) compared three approaches to teaching a set of biology concepts

to tenth graders: student‐led small group discussion, teacher‐led whole‐class discussion, and individual

completion of worksheets. Results showed that both types of discussion led to higher scores on

assessments of the biology vocabulary than did individual exercises, but there was no difference between

the two discussion types. Stahl and Vancil (1986) found similar results with fifth grade students who were

learning a set of meteorology concepts.

These findings suggest that discussion can play a role in the acquisition of domain-specific

academic vocabulary, at least when the words have been explicitly taught as well as integrated into

discussion. Some form of discussion is often one component of interventions to improve vocabulary. For

example, Lesaux et al (2010) implemented small‐group work in addition to whole-class and independent

activities to teach six-grade students a list of academic words, and Snow et al (2009) incorporated

argumentative discourse as a complement to explicit vocabulary instruction. The role small group work

and argumentative discourse play in academic vocabulary acquisition has yet to be clearly established,

however, because in previous studies these practices have been intermingled with other types of

vocabulary instruction.

Social factors in classroom learning.

Social dynamics are important to classroom learning outcomes. According to a comprehensive

review of the literature by Howe and Mercer (2007), children’s social histories, and characteristics such


as popularity, status in the classroom, temperament, and social experience at home and in the community,

impact the quality of interaction in the classroom. Murphy and Faulkner (2000) reported that popular

children were more likely to maintain successful collaboration as compared to unpopular children,

because popular children were more capable of following rules (e.g., turn-taking) and more strategic in

verbal and non-verbal communication such as providing elaborated arguments and monitoring group

members’ facial expressions. Lin et al (2015) found that students’ status in the classroom social network

mediated the effectiveness of collaborative discussions in improving relational thinking. Students

centered in the network provided both more support and more challenges to their peers and played a role

in creating a harmonious and productive experience for all.

Students’ acquisition of academic vocabulary is likely to depend upon the quality of interaction.

Talkativeness, leadership qualities, and social status or, conversely, social insecurity, likely influence

students’ willingness to try new words. Toward a more complete understanding of children’s vocabulary

development, the present study considers both social and cognitive factors that affect peer interaction. To

our knowledge, this study is the first that seeks to understand the influence of social factors on children’s

acquisition of academic vocabulary.

Rationale for the present study

The present study investigated how collaborative group work and whole‐class direct instruction

impact underserved children’s productive use of academic vocabulary, as compared to a control condition

in which students continued regular instruction. As a vehicle for investigating types of instruction, we

developed a curriculum unit, called the Wolf Reintroduction and Management Unit, intended to be

conceptually rich and intellectually stimulating that integrates language arts, science, math and social

studies (Jadallah et al, 2009). Students played the role of officials at the Wolf Management Agency who

are responsible for deciding whether or not a community should be allowed to eradicate a wolf pack

spotted nearby. Three domains of knowledge were introduced to discuss this sensitive policy issue taking

into account potential effects on the local ecosystem, the town’s economy, and public policy. Key terms

were defined as they appeared in the text and were repeated in the margins and in the glossary at the end


of each booklet. In an individual oral transfer task that took the form of an interview, students heard a

statement about the pros and cons of whaling, and then were asked for their own position on whether or

not whaling should be allowed. The transcripts of students’ interviews were searched for both general and

domain-specific academic vocabulary introduced in the Wolf Unit.

This study was motivated by several gaps in current research. First, to our knowledge, the present

study is among the few to investigate the productive use of academic vocabulary. Second, despite the

generally recognized importance of multiple and varied exposure to words, few studies have addressed

how conditions of teaching and learning academic vocabulary words affect the likelihood of students

utilizing the words in contexts different from the one in which they were learned. Third, while the value

of discussion in promoting academic vocabulary is widely appreciated, little research has been dedicated

to determining how and why it works. Lastly, except for facets of language ability, research has yet to

explore how other individual and social characteristics of students influence their uptake and use of

academic vocabulary.

To fill these gaps, this study addresses three research questions: [1] How do collaborative small

groups and whole class direct instruction impact student productive use of general academic vocabulary

and domain-specific vocabulary? Our expectation was that collaborative group work would exceed direct

instruction, as well as the control that continued regular instruction. One basis for this expectation is that

collaborative group work provides more numerous and varied opportunities to use academic vocabulary

words. [2] How does classroom dialogue affect students’ productive use of academic vocabulary? We

expected that frequency and quality of uses of academic vocabulary during classroom talk would mediate

students’ productive use of academic vocabulary in the later transfer task. [3] To what extent do students’

social characteristics influence their use of academic vocabulary words? We anticipated that socially

centered students (and talkative students judged by their peers to be leaders or to have good ideas) would

use more academic vocabulary than socially peripheral students because centered, or high status, students

are more likely to take the initiative to try new words.


Method

Participants

The sample consisted of 460 fifth-grade students enrolled in 36 classrooms in eight public schools

in low-income school districts in central and northern Illinois, who participated in this project across two

academic years. Each year, 18 classrooms were recruited, nine classrooms with a predominant enrollment

of African American students and another nine classrooms with a predominant enrollment of Spanish-

speaking English language learners. Classrooms within triples of classrooms matched on demographic

characteristics and previous academic performance were randomly assigned to one of three intervention

conditions: collaborative group work (CG), direct instruction (DI), or wait-listed control that continued

regular instruction and received the intervention in the following semester after the data were collected.

The 460 participants included 160 CG students, 153 DI students, and 147 Control students,

mainly African American (41%) and Latina/o (49%), approximately balanced by gender (Girl: N=245;

Boy: N=215), average age 10.7 (SD=0.5). Depending on the school from 79% to 99% of the students

were registered for free or reduced lunch. The first language of most of the Latinas and Latinos was

Spanish and, according to a home survey, 84% of them spoke Spanish or a mixture of Spanish and

English with their parents. Thirty-two students (7%) had an Individualized Education Program and

received special services, distributed among conditions as follows: 13 CG students, 12 DI students, 7

Control students.

Procedure

Students in the collaborative group work and direct instruction conditions studied a six-week-long

Wolf Reintroduction and Management Unit that addressed a socio-scientific policy issue faced by an

imaginary community that had requested permission to hire hunters to kill wolves that alarmed many of

its citizens (Jadallah et al., 2009). The unit was divided into three sections, each incorporating an

important perspective on the complicated issue of wolves, to cultivate students’ ability to discern different

aspects of problems and understand interrelationships and trade-offs. The three sections were ecosystem,

economy, and public policy. While killing the wolves may be favored by the majority of residents in the


community (public policy), doing so would alter the food web (ecosystem), which would impact

community businesses (economy). Each section of the unit was explained in an information booklet and

expanded in an activity booklet. Information booklets provided students with essential concepts. Unlike

most readings for middle grade students, the booklets had an argument structure that contrasted opposing

viewpoints.

CG and DI teachers attended parallel two‐day workshops to learn about their assigned

instructional approach to the Wolf Unit. CG teachers learned the theory and research base for

collaborative group work, how to facilitate Collaborative Reasoning discussions, and best practices for

collaborative group work (Gillies, 2007). DI teachers learned the theory and research for whole‐class

teacher directed methods and the best practices for direct instruction (IES, 2007). Teachers saw videos of

the Wolf Unit being taught by the method they were supposed to use.

The intervention encompassed about 22 class sessions over six weeks. In the collaborative group

work condition, the Wolf Unit was implemented in a modified jigsaw format. Students were divided into

three heterogeneous ‘expert’ groups to study one of the three domains: ecosystem, economy or public

policy. At the beginning of the unit, each expert group in the CG condition had a Collaborative Reasoning

discussion in which they talked about their initial opinions on the central question: whether the

community should be permitted to hire professional hunters to kill the wolves. Students were encouraged

to take positions and defend their positions with arguments and evidence in order to make the policy

decision. Then each expert group studied its assigned domain through readings and group activities. After

four weeks, each expert group taught the rest of the class what they had learned by presenting posters.

Following the poster presentations, students were assigned to new groups that contained experts from

each knowledge domain. The new groups had a second Collaborative Reasoning discussion to once again

reflect on the central question.

In the direct instruction condition, the Wolf Unit involved whole‐class teacher‐directed

instruction and individual seatwork, but no small group discussions or other collaborative group work or


poster presentations. Students learned all three knowledge domains through teacher explanation and

teacher‐managed whole‐class discussion. Students read the information booklets and completed the

activity booklets individually at their seats.

In the wait-listed control condition, classrooms did not study the wolf curriculum during the

intervention period but continued regular instruction. Control classrooms had the opportunity to

experience the Wolf Unit in the following semester after the data were collected.

Initial language measures. Pretest assessments included the Gates-MacGinitie reading

comprehension test (MacGinitie, MacGinitie, Maria, & Dreyer, 2000) and a speeded object naming task

(Snodgrass & Vanderwart, 1980) intended to assess children’s basic oral English proficiency. Reading

scores were corrected for guessing (Right – Wrong/3), which improved reliability and predictive validity.

Oral English proficiency was indicated by number of common objects ─ such as bike, car, and bus ─

correctly named per minute.

Initial measures of social characteristics. A sociometric questionnaire elicited peer nominations

and peer ratings to provide assessments of popularity, friendships, social status, talkativeness, quietness,

leadership, and reputation for having good ideas. Children’s talkativeness was calculated by deducting the

number of classmates’ nominations for quietness from the number of nominations for having a lot to say

in class discussions. Three indices of students’ social status, or centrality, were derived from friendship

nominations using social network analysis (Butts, 2008). Indegree centrality represents an individual’s

popularity; it refers to how often a student was nominated by his/her classmates as a friend. Betweenness

centrality represents how often a student was nominated as the common friend of two other unconnected

students. Information centrality represents how far away a student is from every other student in the

friendship network. Peer-liking ratings were summed ratings on a five-point Likert scale of how much

children liked to play with each of the other children in the class. To adjust for differences in class size,

all of the social measures were divided by the number of children in the class.

Whale policy transfer interview. In the individual oral transfer task that followed the Wolf Unit,

each student heard a 386-word statement about the pros and cons of whaling and then was asked to


explain his/her own position on whether whaling should be allowed. The examiner gave standardized

prompts if the student stopped short of providing a complete argument (see Appendix A for more detail).

Student responses to the whale question were audio recorded and transcribed verbatim following the

Systematic Analysis of Language Transcripts (SALT) conventions (Miller & Chapman, 2010).

Due to time and resource constraints, only about 60% of the students were interviewed to present

their opinions on the whaling question. Interviewed first were ‘target students,’ so-called because these

were the students on whom the video camera was trained throughout the unit, selected from each class

with the help of teacher to be a representative sample of the class in terms of gender, ethnicity, academic

performance, and talkativeness. In CG classrooms, target students were the ones who studied the

ecosystem knowledge domain and who worked together every day as a group. In DI classrooms, target

students were a representative sample of the class, selected according to the same criteria as CG target

students, who sat together and were videotaped every day but who did not meet for group activities. In

Control classrooms, there were nominal target students, also selected at the beginning of the study to be a

representative sample of the class in terms of gender, ethnicity, academic performance, and talkativeness,

but who were not videotaped during the period of the intervention and did not work together as a group.

In the week after the intervention was completed, students were pulled out of the class one by one

to be interviewed. The interview was completed one class at a time. Research assistants who conducted

the interview were given a list of randomly ordered names of all the students in the class, with target

students on the top and non-target students filling out the rest of the list. Target students were interviewed

first and then as many non-target students as possible following the order of the list. Eventually, in CG

classrooms, we interviewed 88 ecosystem students, 36 economy students, and 36 public policy. In DI

classrooms, 153 target and non-target students were interviewed while in Control classrooms 147 target

and non-target students were interviewed.

We created a two-level mixed-effects regression model to compare the pretest performance and

social characteristics of students who were interviewed about the whale question and students who were

not. Pretest reading comprehension, object naming, talkativeness, good idea nomination, leader


nomination, and peer liking rating were dependent variables, intervention condition and whether or not

the student received the whale interview were fixed effects. Classroom was entered as a random effect to

account for variance due to teacher or cohort. There was no difference in the object naming of students

who were interviewed and students who were not interviewed, F(1, 725) = 0.35, p = .56. Students who

took the interview had higher pretest reading comprehension scores than those who did not, F(1, 725) =

9.27, p < .01. However, this difference was observed in all three conditions, as indicated by the

nonsignificant condition effect, F(2, 725) = 0.84, p = .43. Students who took the interview were also more

talkative, F(1, 725) = 6.63, p = .010, more likely to be nominated as having good ideas, F(1, 725) =

10.99, p < .01, and more likely to be nominated as leaders, F(1, 725) = 12.35, p < .01; however, these

differences applied to all three conditions. Students who did and did not take the interview were equally

liked by their classmates, F(1, 725) = 0.10, p = .75. Because of time constraints, research assistants were

instructed to skip students on the randomly ordered name list who were absent. Students whose name was

called but were not interviewed either were absent from school on that day or were receiving special

instruction somewhere else in the school as part of their Individualized Education Program. Apparently,

children with high reading scores and children considered to be talkative or to have good ideas or

leadership qualities were less likely to be absent from class.

Identifying academic vocabulary in the Wolf Unit printed materials.

Following Bailey’s (2006) classification of academic vocabulary, student use of general and

domain‐specific academic vocabulary was investigated in this study. A list of 60 domain‐specific words

that convey the core concepts of the Wolf Unit were identified by the authors of the curriculum. There

were 25 Ecosystem words (e.g., food web, predator, species), 17 Economy words (tourism, economy,

ranching), and 18 Public Policy words (advocate, common good, representation). To determine the print

exposure of these domain-specific words for students in the CG expert groups and the DI students, the

Wolf Unit materials for each group were searched to determine the frequency of occurrence each of the

words (see Appendix B for more detail). The inflectional and derivational variants of words were

included in the counts. For example, the exposure frequency of the word reintroduce would include that


of reintroducing, reintroduced, and reintroduction. Expert groups in the CG classrooms had a high rate of

exposure to the domain-specific words in their own domain, but less exposure to the words outside this

domain. Four domain-specific words from the Wolf Unit appeared in the 386-word statement that the

examiner read to the child in the whale transfer task, namely opinion, population, endangered, and

tradition.

The identification of general academic vocabulary (GAV) in the Wolf Unit began with two lists

of academic words. Coxhead’s (2000) Academic Word List contains 570 word families that university

students frequently encounter in textbooks, excluding any of the first 2,000 most frequent English words.

Shoebottom’s (2008) list of 1,040 general academic words was generated from a corpus of words that

second language learners are expected to acquire by their second or third year of intensive English study.

A search for the general vocabulary words from these two lists in the printed Wolf Unit materials netted

250 words that appeared at least once.

Coxhead’s and Shoebottom’s lists are skewed toward the vocabulary needed by college students

and may not capture all of the general academic vocabulary that it would be useful for fifth graders to

know. To see if more words ought to be included, we examined all the low frequency words in the Wolf

Unit materials and found 75 more words that we judged met the criteria for general academic vocabulary.

The first 50 most frequent general academic vocabulary words are presented in Appendix C. The printed

materials in the Wolf Unit were somewhat different for DI students and for different subgroups of CG

students. These variations had to be taken into consideration in estimating the opportunity to learn

domain-specific academic vocabulary. A search of the different versions of the Wolf Unit printed

materials showed that the range of types of general vocabulary words were similar across groups and

conditions (NCG_Ecosystem = 286; NCG_Economy = 272; NCG_Public Policy = 306; NDI = 299), suggesting that

whatever their assignment students had approximately the same opportunity to learn general academic

vocabulary words.


Identifying academic vocabulary in whale interviews.

The whale interview transcripts were searched for the 60 domain-specific words and the 325

general academic words using the text search query and word frequency function of NVivo 10 software.

Of the 325 general academic vocabulary words that appeared in the Wolf Unit printed materials, 88 words

were used by children in their oral whale interview responses. An additional 70 words from Coxhead’s or

Shoebottom’s lists that do not appear in the Wolf Unit printed materials were also identified in the

interviews. We used vocabulary types instead of tokens in the following analyses because types represent

breadth of word knowledge. As the frequency of vocabulary types is a count variable conforming to the

Poisson distribution, mixed-effects Poisson regression models were constructed to examine condition

differences in the use of general or domain-specific academic vocabulary in the whale transfer interviews.

Results

Pretest language and social measures

Separate two-level regression analyses were conducted to check whether there was a difference

between the CG condition, DI condition, and control condition in pretest reading comprehension or

pretest object naming, with gender, ethnicity, as well as intervention condition, as fixed effects.

Classroom was entered as a random factor at the second level to account for teacher or cohort effects. The

results indicated no significant condition difference in reading comprehension, F(2, 420) = 0.76, p = .47,

or object naming, F(2, 420) = 0.69, p = .50. There was a marginal gender difference in reading

comprehension; girls had slightly higher reading scores than boys, F(1, 420) = 3.05, p =.082. Latina/o

children were slower naming common objects than African American children, mean difference = -8.84,

t(420) = -5.60, p < .001. No ethnic difference was found in reading comprehension and no gender

difference was found in object naming.

Two-level regression analyses were performed to evaluate differences in social characteristics

with gender, ethnicity, and intervention condition as fixed effects and classroom as a random variable.

Dependent variables were talkativeness, good idea nominations, leadership nominations, indegree

centrality, betweenness centrality, information centrality, and peer-liking rating. No condition difference


was found in any of these indices. Girls were found to be less talkative than boys, F(1, 420) = 8.18, p

< .01. However, girls were more popular than boys, F(1, 420) = 12.49, p < .001, more likely to be

nominated as leaders, F(1, 420) = 27.25, p < .001, more likely to be nominated as having good ideas, F(1,

420) = 18.11, p < .01, had higher social status as indicated by information centrality, F(1, 420) = 17.44, p

< .001, and were more liked by classmates, F(1, 420) = 32.00, p < .001. African American children were

considered by their peers to be more talkative than Latina/o children, t(420) = 1.98, p = .048.

Uses of academic vocabulary in response to whale question

A total of 158 general academic word types and 31 domain-specific word types were present in

children’s whale policy interviews. Eighty percent of the students used one or more types of general

academic words for a total frequency of 1,537. Forty-four percent of the students used one or more types

of domain-specific words for 1,009 total uses.

Among the 158 general academic vocabulary words occurring in the whale interview, 17 words

were used more than 20 times. These were reason, agree, disagree, affect, instead, fair, spend, argument,

increase, decrease, survive, since, attack, amount, environment, fault, and harm. Domain-specific

vocabulary occurring in the whale interview included 16 ecosystem words, 5 economy words, and 10

public policy words. Twelve words were used more than 10 times in the interviews, including seven

ecosystem words (ecosystem, endangered, extinct, food web, nature, population, species), two economy

words (economy, tourism), and three public policy words (majority, opinion, tradition). Use of ecosystem

words was observed among 64% of CG ecosystem students, 65% of CG economy students, 79% of CG

public policy students, 50% of DI students, and 40% of control students. It would seem, therefore, that

among CG students more use of ecosystem words than economy and public policy words was not due to

more ecosystem students being interviewed.

Table 1 presents means and standard deviations of the occurrence of different types of academic

vocabulary words by intervention condition. Students in collaborative groups showed a higher frequency

of general academic words and a slightly higher frequency of domain-specific words than students who

received direct instruction. Both CG and DI students used more academic words of both types than


control students. Ecosystem words (65%) were most frequently used, economy words (3%) were least

used, and the use of public policy words (35%) was in the middle.

[Insert Table 1 here]

Table 2 presents the zero-order Pearson correlations between use of academic vocabulary in the

whale interview and pretest language and social measures. Both reading comprehension and object

naming were significantly correlated with the use of general academic words and domain-specific words,

as were several of the social measures.


Factors predicting use of general academic vocabulary. Mixed-effects Poisson regression

models were constructed for the occurrence of GAV words in the whale interview. Fixed effects were

gender, ethnicity, and intervention condition. Classroom was entered as a random variable. Six covariates,

including reading comprehension, object naming, talkativeness, peer-liking rating, leadership

nominations, and information centrality were entered sequentially into the analysis. Gender and ethnicity

were dropped during model selection because neither of them was ever significant, ps >.30. Indegree

centrality and betweenness centrality were not included because of low correlations with use of academic

words and overlap with information centrality. Good idea nomination was included initially but later

dropped because it overlapped with leadership nominations.


The mixed-effects Poisson regression models for the use of GAV are presented in Table 3. The

model selection process started with an empty model that included only the random effect. The empty

model showed significant individual differences in the use of GAV words, after controlling for classroom

variance, intercept = 0.89, t(36) = 15.05, p < .001. Reading comprehension, a measure of students’

receptive language ability, was entered in Model 1, and predicted use of GAV, F(1, 422) = 15.66, p

< .001. Object naming, which is considered to represent students’ basic oral English proficiency, was

entered in Model 2 and also predicted GAV, F(1, 421) = 9.46, p < .01. Models 3 through 6 accounted for

social characteristics. There were significant effects for talkativeness, F(1, 420) = 6.70, p = .01; the peer-


liking rating of popularity among classmates, F(1, 419) = 9.90, p < .01; leadership nominations, F(1, 418)

= 4.31, p = .039; and information centrality, a measure of students’ status in the classroom social network

that takes into account both direct and indirect ties with others, F(1, 417) = 6.29, p = .013.

After controlling for initial language ability, talkativeness, and social characteristics, we

incorporated intervention condition in Model 7 and found a significant effect, F(2, 417) = 11.73, p < .001.

Pair-wise comparisons showed that CG students used more GAV words than DI students, mean

difference Mdiff = 0.58, t(417) = 2.26, p = .024, and DI students used more GAV words than control

students, Mdiff = 0.64, t(417) = 2.63, p < .01.

To summarize, students who interacted in collaborative groups productively used more general

academic vocabulary in the whale transfer task than students who experienced direct instruction,

controlling for initial language ability, talkativeness, and social characteristics. Both CG and DI students

produced more GAV words than uninstructed control students. Students who were better readers and had

better oral English, who were more talkative, recognized by peers for leadership, or enjoyed higher social

status were more likely to use GAV words in the whale transfer task. The only factor that showed a

negative association with GAV words was the peer-liking measure; students who were more liked by

classmates tended to use fewer general academic words. The foregoing pattern of results remained the

same when we analyzed the total occurrence of academic words including both general and domain-

specific academic vocabulary.

Factors predicting use of domain-specific vocabulary. A mixed-effects Poisson regression

model was constructed for domain-specific vocabulary with gender, ethnicity, and intervention condition

as fixed effects, classroom as a random effect, and the two initial language ability measures and four

social measures as covariates. Both CG and DI students significantly outperformed control students, Mdiff

(CG vs. Control) = 0.79, t(414) = 4.23, p < .001 and Mdiff (DI vs. Control) = 0.67, t(414) = 3.85, p <.001.

However, no difference was observed between the CG and DI conditions, t(414) = 0.35, p = .73. Girls

used more domain-specific vocabulary than boys, F(1, 414) = 5.04, p = .025. Reading comprehension,

object naming, talkativeness, and leadership nominations were significant predictors of use of domain-


specific vocabulary, ps < .05. Peer-liking ratings showed a marginal negative effect, F(1, 414) = 3.28, p

= .071. Social status (information centrality) did not predict the use of domain-specific vocabulary, F(1,

414) = 0.76, p = .38. No difference was found between African American and Latina/o students in use of

domain-specific academic words, F(2, 414) = 1.65, p = .19.

Collaborative group work, although more effective in improving students’ use of general

academic vocabulary, was no better than direct instruction in improving students’ domain-specific

vocabulary. However, students in the CG condition thoroughly studied only one domain of knowledge,

while DI students studied all three domains. CG ‘experts’ might use more words from their own domain

of expertise as compared with words from the two less studied domains. Ordered logistic regression

analyses were conducted to investigate condition differences in use of vocabulary specific to the

ecosystem, economy, and public policy domains. The ordinal variable for use of ecosystem words

contained four categories: zero frequency (n = 219), one type (n = 148), two types (n = 51), and three or

more types (n = 42). There were three categories of uses of public policy words: zero frequency (n =281),

one type (n = 143), and two or more types (n = 36). The proportional odds assumption held for both

analyses. The control condition was used as the reference. After controlling for gender, ethnicity, reading

comprehension, object naming, talkativeness, leader nomination, information centrality, and peer-liking

ratings, the condition effect was significant in the use of ecosystem words, χ2 (4, N = 460) = 19.64, p

<.01, and the use of public policy words, χ2 (4, N = 460) = 16.09, p < .01. No analysis of economy words

was performed due to their rare occurrence (4%); among the 460 students, six CG students and 12 DI

students used one economy word.

Students in the public policy group had the highest chance of using ecosystem words (odds ratio

Public policy/Control = 3.00, 95% CI [1.49, 6.07]), as compared to the economy group (odds ratio

Economy/Control = 2.99, 95% CI [1.48, 6.06]), the ecosystem group (odds ratio Ecosystem/Control =

2.51, 95% CI [1.49, 4.24]), or the DI students (odds ratio DI/Control = 2.08, 95% CI [1.31, 3.30]). Public

policy students also had the highest chance of using public policy words (odds ratio Public policy/Control

= 4.07, 95% CI [1.92, 8.61]), as compared to the DI students (odds ratio DI/Control = 2.11, 95% CI [1.29,


3.47]). However, both ecosystem and economy students had a lower chance of using public policy words

than DI students, odds ratio Ecosystem/Control = 1.63, 95% CI [0.92, 2.87] and odds ratio

Economy/Control = 1.65, 95% CI [0.76, 3.60].

To summarize, students in the three CG expert groups all had a greater likelihood of using

ecosystem words than the DI students, and students in public policy groups also used more public policy

words than the DI students. Bear in mind, though, that when the distinction among expert groups is

collapsed and all of the domain-specific words are aggregated, there is no overall difference between CG

and DI students.

Influence of classroom dialogue on productive use of general academic vocabulary words

We hypothesized that the use of GAV words during the Wolf Unit would play a part in

vocabulary growth and would favor collaborative groups as compared to direct instruction. A simple line

of reasoning to inform this hypothesis is that collaborative group work provides more opportunities to

employ GAV words during discussions and other group activities. This, in turn, enhances their ability to

productively use these words in the whale transfer interview.

To examine this hypothesis, we performed a mediation analysis at the word level. The analysis

encompassed the 158 general academic words used by one or more student during the whale interview.

Control GAV is defined as the proportion of control students (N = 147) who used each of the 158 GAV

words in response to the whale question; since control students did not study the Wolf Unit, their use of

the words provides the baseline for GAV words already known by a fifth grader from a low-income

minority family that were accessible to this fifth grader while trying to respond to the whale question.

Wolf Print GAV is the print exposure frequency of each of the array of 158 GAV words in the curriculum

materials that CG and DI students were supposed to read during the Wolf Unit. The outcome variable was

Whale GAV, the proportion of CG students and DI students who used each general academic word in the

whale response. Intervention Condition (CG = 1, DI = 0) was the explanatory variable of major interest.

Wolf Dialogue GAV, the frequency of student use of each of the array of general academic words in

classroom talk throughout the Wolf Unit, was the candidate mediator variable.


Wolf Dialogue GAV was based on a search for the 158 general academic vocabulary words in

transcripts of 146 four-minute excerpts (total duration = 594 minutes) sampled from 500 hours of video

recorded during Wolf Unit lessons in the CG and DI classrooms. The four-minute excerpts were selected

according to a stratified random sampling plan. One excerpt was sampled from each of six important

lessons (seven in one case) in each CG classroom (N = 12) drawn from the following: introduction to the

Wolf Unit, first Collaborative Reasoning discussion, wolves in the United States, wolves and the

ecosystem, a poster presentation of major concepts in the ecosystem domain, and the second

Collaborative Reasoning discussion. Likewise, one excerpt was sampled from each of six important

lessons (seven in one case) in each DI classroom (N = 12) to cover the introduction to the Wolf Unit,

wolves in the United States, wolves and the ecosystem, wolves and the economy, wolves and public

policy, and review of major concepts in the Wolf Unit. The sampled lessons were spaced at

approximately equal intervals across the unit and usually occurred on Tuesdays. The first few and last few

minutes of each lesson were trimmed because non-instructional activities are likely at these times (passing

out materials, lining up for lunch). The four-minute excerpt was selected at random from the remainder of

the lesson with the constraint that in any one classroom three of the six excerpts came from the first half

of a sampled lesson and three came from the second half of a sampled lesson.

Separate counts of general academic vocabulary words were obtained for student uses and teacher

uses. The average rate of student use of GAV words per four-minute excerpt was 4.63 in CG classrooms

and 2.15 in DI classrooms. In contrast, the average rate of teacher use of GAV words per four-minute

excerpt was 1.34 in CG classrooms but 3.64 in DI classrooms. In the principal mediation analysis

described below, Wolf Dialogue GAV refers to student uses of GAV words during the wolf management

unit.

The mediation analysis involved a subset of 88 of the students from the CG condition and all

153 students from the DI condition who were interviewed about whales. The analysis had to be limited to

the subset of CG students whose classroom dialogue had been video recorded. These were the students in

ecosystem groups. Their dialogue could have been quite different from the dialogue of students in


economy groups or public policy groups who were not video recorded. In DI classrooms, the video

recorded dialogue was representative of all classroom dialogue so all 153 DI students who were

interviewed about the whale question could be included in the analysis.

Since the outcome variable (Whale GAV) and the mediation variable (Wolf Dialogue GAV)

followed a Poisson distribution, we constructed three Poisson regression models to determine the

coefficients that define direct and indirect effects. The three Poisson models examined [1] the effect of

intervention condition on Whale GAV controlling for Control GAV and Wolf Print GAV (see Figure

1A), [2] the effect of intervention condition on Wolf Dialogue GAV after controlling for Wolf Print GAV

(see Figure 1B), and [3] the effect of intervention condition on Whale GAV after entering the mediator

variable Wolf Dialogue GAV and the two covariates, Control GAV and Wolf Print GAV (see Figure 1B).

[Insert Figure 1 here]

As can be seen in Figure 1B, there are two indirect paths involving Wolf Dialogue GAV. One is

Intervention Condition Wolf Dialogue GAV Whale GAV and the other is Wolf Print GAV Wolf

Dialogue GAV Whale GAV. The possible mediating effect of Wolf Dialogue GAV via the path

Intervention Condition Wolf Dialogue GAV Whale GAV was tested first. The analytical method we

employed was developed by Iacobucci (2012) for cases in which the independent variable (X), mediator

variable (M), and/or outcome variable (Y) are categorical. In line with Baron and Kenny (1986),

coefficient a represents the direct effect of X on M, b represents the direct effect of M on Y when X and

M are both in the model, c represents the direct effect of X on Y without M in the model, and c′ the effect

of X on Y with M in the model. The standard errors of a and b are denoted as sa and sa and can be

obtained from the regression analyses. The test statistic Zmediation was defined by Iacobucci (2012) as

𝑍𝑚𝑒𝑑𝑖𝑎𝑡𝑖𝑜𝑛 =

𝑎𝑠𝑎

×𝑏𝑠𝑏

√(𝑎𝑠𝑎

)2 + (𝑏𝑠𝑏

)2 + 1

Poisson Model 1 establishes that intervention condition significantly predicted Whale GAV, c =

0.28, χ2 (N = 316) = 4.90, p = .027, after controlling for Control GAV and Wolf Print GAV. Poisson


Model 2 indicates a significant effect of intervention condition on Wolf Dialogue GAV, a = 0.66, χ2 (N=

316) = 30.43, p < .001. In Model 3, the effect of intervention condition on Whale GAV is not significant

after incorporating the mediator variable Wolf Dialogue GAV in the model, c′ = 0.07, χ2 (N = 316) =

0.23, p = .64. Instead, in Model 3 there is a significant effect of Wolf Dialogue GAV on Whale GAV, b =

0.15, χ2 (N = 316) = 8.17, p < .01. Based on Iacobucci’s (2012) method, there is a mediation effect of

Wolf Dialogue GAV, Zmediation = 2.51, p < .01. Therefore, the analysis indicates that Wolf Dialogue GAV

mediated the effect of intervention condition on Whale GAV.

We also tested whether there was a mediation effect from teachers’ use of GAV words in

classroom dialogue during the wolf management unit. The results indicated that teachers’ use of GAV

words did not predict students’ use of GAV words in the whale interview and did not mediate the effect

of instructional condition on Whale GAV: a = -0.50, χ2 (N= 316) = 12.75, p < .001; b = 0.05, χ2 (N = 316)

= 0.85, p = .36; c = 0.28, χ2 (N = 316) = 4.90, p = .027; c′ = 0.31, χ2 (N = 316) = 5.63, p = .018; and

Zmediation = -0.86, p = .19. An analysis of total GAV words in classroom dialogue that included both

teacher and student uses of words yielded results similar to the analysis involving only student uses, but

the mediation effect of Wolf Dialogue GAV was not as strong.

The second possible mediating effect of Wolf Dialogue GAV is via the path Wolf Print GAV

Wolf Dialogue GAV Whale GAV. Results show that Wolf Dialogue GAV partially mediated the effect

of Wolf Print GAV on Whale GAV, Zmediation = 2.75, p < .01. To summarize, the frequency with which

students used general academic vocabulary in classroom dialogue partially mediated the effect of

intervention condition and the influence of the printed wolf management curriculum on use of general

academic vocabulary words in the whale interview.

Discussion

An important finding from the present research is that students who studied wolf management for

six-weeks, addressing the question of whether a community should be permitted to eradicate a pack of

wolves, exceeded uninstructed control students in productive use of academic vocabulary in an open-

ended transfer task that required students to justify a position on whether or not whaling should be


allowed. Students who experienced the Wolf Unit exceeded control students in use of both general

academic vocabulary words, such as increase and protect, and domain-specific vocabulary words, or

technical terms, such as population and extinct.

Few if any previous studies have demonstrated that instruction can have a broad impact on

students’ ability to productively use academic words, where by productive use we mean the ability to say

or write the words spontaneously, in contrast to recognizing the words, selecting their definitions from a

set of options, or supplying them in response to focused prompts. That students had active control of a

cross section of the academic vocabulary words they encountered in the Wolf Unit is further suggested by

the fact that they were able to use the words in response to the whale question, which differed in many

surface aspects from the wolf question they had studied.

Our general explanation for why students used more academic vocabulary words in the oral

transfer task after completing the Wolf Unit is that the unit provided for multiple encounters with words

in meaningful contexts. Students were exposed to a wide variety of academic words as they studied

different topics related to the controversial policy issue about wolves. In order to make a thoughtful

decision about the policy issue, students needed to organize information about the different topics into

arguments. Students in collaborative groups acquired academic words useful for organizing and

expressing information about wolves in give and take with peers, whereas students who received direct

instruction acquired the words by following the teacher’s explanations and using the words in response to

the teacher’s questions.

A second important finding of this research is that students who experienced the Wolf Unit via

Collaborative Groups (CG) used more general academic vocabulary (GAV) words on the oral transfer

task than students who experienced the Wolf Unit via Direct Instruction (DI). Classroom dialogue during

collaborative peer interaction and teacher-led instruction differs in many ways (Chinn, Anderson, &

Waggoner, 2001) and it is here that we looked for an explanation for the greater use of GAV words by

CG than DI students in response to the whale question. A simple explanation for the difference is that

GAV words were used more frequently in the classroom dialogue in CG classrooms.


This simple explanation was evaluated in the analysis summarized in Figure 1. After controlling

for prior knowledge (Control GAV) of an array of 158 general academic vocabulary words that appeared

in the whale interviews, CG students were significantly more likely than DI students to use the array of

words in response to the whale question (Whale GAV). But, after incorporating the frequency of students’

use of the words in classroom talk in CG and DI classrooms (Wolf Dialogue GAV) in the model, the

direct effect of instructional condition was no longer significant. In this model, the effect of instructional

condition was indirect. Instructional condition predicted student use of GAV words in classroom dialogue

and student use of the words in classroom dialogue predicted their use in the whale interview. Thus, the

effect of collaborative group work appears to have been mediated at least in part by the greater frequency

of academic vocabulary words in students’ talk in collaborative group classrooms.

Beyond frequency of word use, another factor that could have contributed to the advantage of CG

students is heightened engagement. Wu and her colleagues (2013) reported that collaborative peer

discussions result in greater student interest and engagement than conventional teacher-led discussions.

Students motivated to actively join a discussion may thereby use some academic vocabulary words and

more closely follow others’ use of academic words.

More opportunity to actively use academic vocabulary words in dialogue figures to be a major

reason for CG students’ greater uptake and later use of the words. Opportunities for speaking are

expanded when a class is divided into small groups. Time must be divided among all the students in

whole-class discussions whereas time is divided among fewer students in small-group discussions. During

teacher-led discussions, time must be split between the teacher and the students, and teachers often take

much of the time and may express more than half the words that are spoken during discussions (Cazden,

2001). In collaborative small-group discussions among peers, teacher time is nil and almost all of the time

is available for student speaking turns. Dispensing with the rigmarole of hand raising and teacher

nomination for speaking turns saves additional time during peer-to-peer discussions and enables a focus

on ideas. It is not surprising, therefore, that Chinn, Anderson, and Waggoner (2001) found that student

words per minute was nearly twice as high in peer-managed discussions in small groups as compared to


teacher-led discussions in the same small groups. In the present study, rate of use of general academic

vocabulary words during classroom dialogue by CG students was over twice as high as the rate of use by

DI students. Conversely, the rate of use was nearly three times higher among DI teachers than CG

teachers.

Richer classroom dialogue is a probable factor contributing to the advantage of CG students as

compared to DI students in use of GAV words in the whale interview. A body of evidence establishes that

improved comprehension, learning, and problem solving are associated with high quality classroom

discussion during which students’ provide explanations (as opposed to merely listening to them),

elaborate ideas by linking them with prior knowledge, predict the consequences of different courses of

action, draw inferences that connect different parts of texts, construct analogies between real and

imagined situations, consider alternative explanations, support ideas with evidence, build on one

another’s ideas to co-construct explanations, and critique one another’s ideas (Nystrand et al., 1997;

Murphy et al., 2009; Resnick & Schantz, 2015; Reznitskaya et al., 2009). The link to vocabulary

acquisition comes from a large study by Lawrence, Crosson, Paré-Blagoev, and Snow (2015)

encompassing over 1,500 middle school students from 28 schools. That study evaluated a vocabulary

instruction program called Word Generation, in which students read, discuss, and write about

controversial topics using a list of target academic words. The results showed that the Word Generation

program improved the quality of discussion in a range of classes including math, science, language arts,

and social studies. The program also led to modest gains in academic vocabulary. These gains were

mediated in part by quality of classroom discussion.

A low-inference indicator of the quality of discussions in CG and DI classrooms is use of

coordinating conjunctions. Morris and his colleagues (2013) examined the frequency of use of the

conjunctions because, so, if, then, and, and but during the four-minute excerpts sampled from the

classroom dialogue in the CG and DI classrooms enrolled in the present study. CG students’ rate of use of

the conjunctions was four times higher than the rate of use by DI students. In contrast DI teachers used

the conjunctions at twice the rate of CG teachers. Thus, CG students had the experience of expressing


elaborated and connected ideas during collaborative group work. DI students, in contrast, depended on

teachers to initiate ideas and make connections.

High quality discussion may lead to better lexical representations. According to the Lexical

Quality Hypothesis (Perfetti & Hart, 2002), a good lexical representation contains sufficient knowledge of

orthographic, phonological, and semantic properties of words for people to retrieve the words rapidly and

flexibly. Students who experience the Wolf Unit may develop high quality lexical representations of

academic words that entail a network of connections that link the words to many contexts. Students in

collaborative groups, especially, may develop dense and integrated networks through constantly

presenting claims and responding to classmates’ claims and challenges during collaborative discussions.

The more dense and integrated the network the more likely words are to ‘come to mind’ in new contexts

such as the whale interview.

The wolf management curriculum is encapsulated in printed materials that provide information,

explain concepts, and introduce perspectives on the relationships between wolves and the ecosystem,

wolves and the economy, and wolves and public policy. However, words lying lifeless on a page do not

teach. The curriculum is brought to life in classroom dialogue. The analysis summarized in Figure 1

shows that the frequency of general academic vocabulary words in printed Wolf Unit curriculum

materials (Wolf Print GAV) influenced the frequency of these words in classroom dialogue (Wolf

Dialogue GAV) which in turn influenced whether students would use the words in response to the

question about whaling (Whale GAV). Therefore, in other words, the influence of the printed curriculum

on use of academic vocabulary in the whale interview was mediated by classroom dialogue.

On average, CG students used two more academic word types in the whale interview than control

students, while DI students used about one-and-a-third more types than control students. These perhaps

seem like minor improvements, but it should be stressed again that the children were from underserved

communities and more than half were English language learners, and thus probably most had little

exposure to academic language. The words observed in any finite sample of language are a fraction of the

lexicon from which the words were drawn. For every additional word observed in the language sample it


must be inferred that there were other words that were added to the lexicon that remained unobserved (see

Carroll, 1971; Kojima & Yamashita, 2014). If a six-week long unit can result in one or two more

academic words appearing in a limited sample of speech, other words not observed must also have been

acquired and the long-term yield of new words from this kind of instruction could be substantial.

Talkativeness, nominations for leadership, and position in the classroom social network

(information centrality) were positively related to use of general academic vocabulary words in the whale

transfer interview, while peer-liking showed a negative relationship. These findings provide tantalizing

clues to what must have been the social process that gave rise to academic vocabulary knowledge. We

surmise that talkative, socially-centered children took the lead in using academic vocabulary words and

trying to figure out their meanings. In a study of peer influences during collaborative reasoning, Lin and

her colleagues (2015) concluded that, “students at the center of the classroom friendship network play an

influential role in creating a stimulating and friendly discussion environment in which everyone has the

opportunity to make contributions (p. 94).” The negative relationship between peer-liking and use of

academic vocabulary apparently means that, aside from children who are well-liked because of their

intellectual and social leadership, the remaining children who are well-liked tend to be averse to academic

talk and to shy away from using academic vocabulary words.

CG students and the DI students did not differ overall in use of domain-specific vocabulary in the

whale interview. However, when the fact that ‘expert’ groups within the CG condition concentrated on

different domains of knowledge is taken into consideration, a somewhat different picture emerges.

Students in ecosystem expert groups more often used ecosystem words than DI students, and students in

public policy expert groups more often used public policy words than DI students. Results were flat for

economy words since these were seldom used in the whale interview. Interestingly, even students in the

public policy and economy expert groups used significantly more ecosystem words than DI students.

With the caveat that we have not yet examined exactly which words were taught, and how they were

taught, the findings so far seem to give reason to doubt the received wisdom that technical vocabulary

cannot be learned unless explicitly taught.


A limitation of this study is that there was no pre-intervention assessment of children’s ability to

productively use academic vocabulary in an open-ended speaking task. A pre-intervention assessment

would have provided further assurance that groups were comparable and allowed more sensitive tests of

intervention effects. A second limitation is that untangling the effects of instructional conditions on

domain-specific vocabulary acquisition was tricky because of the complicated instructional design. CG

students studied one domain of knowledge intensively, while DI students were exposed to all three

domains. A third limitation was that, because of constraints on time and resources, only 60% of the

students could be interviewed about the whale question and the analysis of classroom dialogue was

limited to 2% of the available lesson video.

Although productive vocabulary has long been distinguished from receptive vocabulary, “the idea

of productive vocabulary remains a fundamentally elusive one. The main reason for this is that it has

proved surprisingly difficult to develop simple and elegant tests of productive vocabulary. . .” (Meara &

Alcoy, 2010, p. 222). Laufer and Nation (1999) developed a productive vocabulary assessment that may

meet the requirement of being simple and elegant. It entails completing words in sentences. The first

letters of each incomplete word are provided. The letter cues and cues from sentence meaning converge

on one and only one word, so while the method assesses productive vocabulary, it does so only at a

minimal level.

As far as we know, the present study is the first to demonstrate acquisition and spontaneous

productive use of academic vocabulary in an open-ended oral transfer task. Most research on vocabulary

instruction employs tests of receptive knowledge of words explicitly taught. A standardized vocabulary

test containing mostly words beyond those taught may be given in addition, in the hope that students have

learned something generalizable or transferable. Usually this is a vain hope. Vocabulary instruction

seldom improves performance on standardized vocabulary tests, perhaps because such tests do not fully

reflect students' vocabulary competence (cf. Pearson, Hiebert, & Kamil, 2007). Findings of the present

study warrant renewed optimism that instruction can have a broad impact on students’ ability to

understand and productively use academic vocabulary words.


References

Anderson, R. C., & Freebody, P. (1981). Vocabulary knowledge. In J. Guthrie (ed.), Comprehension and

Teaching: Research Reviews (pp. 77‐117). Newark, DE: International Reading Association.

Anderson, R. C., Chinn, C., Waggoner, M., & Nguyen-Jahiel, K. (1998). Intellectually stimulating story

discussions. In J. Osborn & F. Lehr (Eds.), Literacy for all: Issues in teaching and learning. (pp.

170–186). New York: Guilford.

Anderson, R. C., Nguyen-Jahiel, K., McNurlen, B., Archodidou, A., Kim, S.-Y., Reznitskaya, A., …

Gilbert, L. (2001). The snowball phenomenon: Spread of ways of talking and ways of thinking

across groups of children. Cognition and Instruction, 19, 1-46.

Applebee, A. N., Langer, J. A., Nystrand, M., & Gamoran, A. (2003). Discussion‐based approaches to

developing understanding. American Educational Research Journal, 40(3), 685–730.

August, D., Carlo, M., Dressler, C., & Snow, C. (2005). The critical role of vocabulary development for

English language learners. Learning Disabilities Research & Practice, 20, 50–57.

August, D., Branum‐Martin, L., Cardenas‐Hagan, E., and Francis, D.J. (2009). The impact of an

instructional intervention on the science and language learning of middle grade English language

learners. Journal of Research on Educational Effectiveness, 2(4), 345–376.

Baron, R. M., & Kenny, D. A. (1986). The moderator–mediator variable distinction in social

psychological research: Conceptual, strategic, and statistical considerations. Journal of

Personality and Social Psychology, 51(6), 1173–1182.

Barron, R. F., & Melnick, R. (1973). The effects of discussion upon learning vocabulary meanings and

relationships in tenth grade biology. In H. L. Herber & R. F. Barron (Eds.), Research in reading

in the content areas: Second year report (pp. 46‐52). Syracuse, NY: Syracuse University,

Reading and Language Arts Center.

Bailey, A. L. (2006) (Ed). Language demands of school: Putting academic English to the test. New

Haven, CT: Yale University Press.

Beck, I. L., McKeown, M.G., & Kucan, L. (2002). Bringing words to life: Robust vocabulary instruction.

New York: Guilford.

Bravo, M. A., & Cervetti, G. N. (2008). Teaching vocabulary through text and experience in content

areas. In A. E. Farstrup & S. J. Samuels (Eds.), What research has to say about vocabulary

instruction (pp. 130‐149). Newark, DE: International Reading Association.

Butts, C. T. (2008). Social network analysis with sna. Journal of Statistical Software, 24, 1-51.

Carlisle, J. F., Fleming, J. E., & Gudbrandsen, B. (2000). Incidental word learning in science classes.

Contemporary Educational Psychology, 25(2), 184–211.

Carlo, M., August, D., McLaughlin, B., Snow, C., Dressler, C., Lippman, D., Lively, T., & White, C.

(2004). Closing the gap: Addressing the vocabulary needs of English language learners in

bilingual and mainstream classrooms. Reading Research Quarterly, 39(2), 188–215.


Carroll, J. B. (1971). Statistical analysis of the corpus. In J. B. Carroll, P. Davies, and B. Richman,

American Heritage Word Frequency Book. New York: Houghton Mifflin.

Clark, A. M., Anderson, R. C., Archodidou, A., Nguyen‐Jahiel, K., Kuo, L.‐J., & Kim, I. (2003).

Collaborative Reasoning: Expanding ways for children to talk and think in the classroom.

Educational Psychology Review, 15, 181‐198.

Corson, D. (1997). The learning and use of academic English words. Language Learning, 47, 671–718.

Coxhead, A. (2000). A new academic word list. TESOL quarterly, 213-238.

Ellis, R. (2005). Instructed second language acquisition: A literature review. Wellington, New Zealand:

Ministry of Education.

Garcia, G. E. (1991). Factors influencing the English reading test performance of Spanish‐speaking

Hispanic students. Reading Research Quarterly, 26, 371–392.

Gillies, R. M. (2007). Cooperative Learning: Integrating Theory and Practice. Thousand Oaks, CA: Sage

Publications.

Hiebert, E.H., & Lubliner, S. (2008). The nature, learning, and instruction of general academic

vocabulary. In A. E. Farstrup & S. J. Samuels (Eds.), What research has to say about vocabulary

instruction (pp. 106‐129). Newark, DE: International Reading Association.

Hoff, E. (2006). How social contexts support and shape language development. Developmental Review,

26(1), 55-88.

Howe, C. & Mercer, N. (2007) Children’s Social Development, Peer Interaction and Classroom Learning

(Primary Review Research Survey 2/1b), Cambridge: University of Cambridge.

Iacobucci, D. (2012). Mediation analysis and categorical variables: The final frontier. Journal of

Consumer Psychology, 22, 582-594.

Institute of Educational Sciences (2007). Organizing Instruction and Study to Improve Student Learning.

Washington, DC: IES.

Jadallah, M., Miller, B., Anderson, R. C., Nguyen-Jahiel, K., Archodidou, A., Zhang, J., & Grabow, K.

(2009). Collaborative Reasoning about a science and public policy issue. In M. McKeown and L.

Kucan (Eds.) Bringing reading research to life: Essays in honor of Isabel L. Beck. New York:

Guilford Press. (pp. 170-193).

Keller, K., Troesch, L. M., & Grob, A. (2015). A large receptive-expressive gap in bilingual children.

Frontiers in Psychology, 6,

Kelley, J. G., Lesaux, N. K., Kieffer, M. J., & Faller, S. E. (2010). Effective academic vocabulary

instruction in the urban middle school. The Reading Teacher, 64(1), 5-14.

Kojima, M., & Yamashita, J. (2014). Reliability of lexical richness measures based on word lists in short

second language productions. System, 42, 23-33.

Laufer, B., & Nation, P. (1995). Vocabulary size and use: Lexical richness in L2 written production.

Applied linguistics, 16(3), 307-322.

Laufer, B., & Nation, P. (1999). A vocabulary size test of controlled productive ability. Language

Testing, 16, 33–51.


Lawrence, J. F., Crosson, A. C., Paré-Blagoev, E. J., & Snow, C. E. (2015). Word Generation

Randomized Trial: Discussion Mediates the Impact of Program Treatment on Academic Word

Learning. American Educational Research Journal, 52, 750-786.

Lesaux, N. K., Kieffer, M. J., Faller, S. E., & Kelley, J. G. (2010). The effectiveness and ease of

implementation of an academic vocabulary intervention for linguistically diverse students in

urban middle schools. Reading Research Quarterly, 45(2), 196‐228.

Lin, T.-J., Anderson, R. C., Hummel, J. E., Jadallah, M., Miller, B. W., Nguyen-Jahiel, K., … Dong, T.

(2012). Children’s use of analogy during Collaborative Reasoning. Child Development, 83, 1429-

1443.

Lin, T.-J., Anderson, R. C., Jadallah, M., Nguyen-Jahiel, K., Kim, I. H., Kuo, L. J., ... & Li, Y. (2015).

Social influences on children's development of relational thinking during small-group

discussions. Contemporary Educational Psychology, 41, 83-97.

Ma, S., Anderson, R. C., Lin, T.-J., Zhang, J., Morris, J. A., Nguyen-Jahiel, ... Latawiec, B. (2015).

Instructional influences on English language learners’ storytelling. Champaign, IL: Center for

the Study of Reading

Marulis, L. M., & Neuman, S. B. (2010). The effects of vocabulary intervention on young children’s

word learning: A meta-analysis. Review of Educational Research, 80(3), 300-335.

MacGinitie, W. H., MacGinitie, R. K., Maria, K., & Dreyer, L.G. (2000). Gates-MacGinitie Reading

Tests (4th ed.), Level 4, Form S. Itasca, IL: Riverside Publishing.

McKeown, M. G., Beck, I. L., Omanson, R. C., & Pople, M. T. (1985). Some effects of the nature and

frequency of vocabulary instruction on the knowledge and use of words. Reading Research

Quarterly, 20, 522–535.

Meara, P. M., Alcoy, J. E. O. (2010). Words as species: An alternative approach to estimating

productive vocabulary size. Reading in a Foreign Language, 22(1), 222-236.

Miller, J. F., & Chapman, R. S. (2010). Systematic Analysis of Language Transcripts [computer

software]. Madison, WI: Language Analysis Laboratory, Waisman Center, University of

Wisconsin.

Mol, S. E., Bus, A. G., & de Jong, M. T. (2009). Interactive book reading in early education: A tool to

stimulate print knowledge as well as oral language. Review of Educational Research, 79(2), 979-

1007.

Morris, J., Miller, B., Anderson, R. C., Lin, T.-J., Nguyen-Jahiel, K., Sun., J., … Wu., X. (2013).

Instructional discourse and argumentative writing. Champaign, IL: Center for the Study of

Reading, University of Illinois.

Murphy, S., & Faulkner, D. (2000). Learning to collaborate: Can young children develop better

communication strategies through collaboration with a more popular peer. European Journal of

Psychology of Education, 15(4), 389-404.

Nagy, W.E. (1997). On the role of context in first‐ and second-language vocabulary learning. In N.

Schmitt & M. McCarthy (Eds.), Vocabulary: Description, acquisition and pedagogy (pp. 64–83).

Cambridge, UK: Cambridge University Press.


Nagy, W. E., & Anderson, R. C. (1984). How many words are there in printed school English? Reading

Research Quarterly, 19, 304‐330.

Nagy, W. E., Anderson, R. C., & Herman, P. A. (1987). Learning word meanings from context during

normal reading. American Educational Research Journal, 24, 237–270.

Nagy, W. E., & Scott, J. (2000). Vocabulary processes. In M. Kamil, P. Mossenthal, P. Pearson, & R.

Barr (Eds.), Handbook of reading research (Vol. 3, pp. 269–284). Mahwah, NJ: Erlbaum.

Newcomb, A. F., Bukowski, W. M., & Pattee, L. (1993). Children's peer relations: a meta-analytic review

of popular, rejected, neglected, controversial, and average sociometric status. Psychological

bulletin, 113(1), 99.

NICHD. (2000).The report of the National Reading Panel. Teaching children to read: An evidence‐based

assessment of the scientific literature on reading and its implications for reading instruction.

Washington, DC: U.S. Government Printing Office.

Nystrand, M., Gamoran, A., Kachur, R., & Pendergast, C. (1997). Opening dialogue: Understanding the

dynamics of language and learning in the English classroom. New York: Teachers College Press.

Oxford, R. L. (1997). Cooperative learning, collaborative learning, and interaction: Three communicative

strands in the language classroom. The Modern Language Journal, 81(4), 443-456.

Pearson, P. D., Hiebert, E. H., & Kamil, M. L. (2007). Vocabulary assessment: What we know and what

we need to learn. Reading Research Quarterly, 42(2), 282-296.

Resnick, L.B., & Schantz. F. (2015). Talking to learn: The promise and challenge of dialogic teaching. In

L. B. Resnick, C. S. C. Asterhan, and S. N. Clarke (Eds.). Socializing intelligence through

academic talk and dialogue. Washington DC: American Educational Research Association

Reznitskaya, A., Kuo, L., Clark, A., Miller, B., Jadallah, M., Anderson, R. C., & Nguyen-Jahiel, K.

(2009). Collaborative reasoning: A dialogic approach to group discussions. Cambridge Journal of

Education, 39, 29-48. doi:10.1080/03057640802701952

Salahu‐Din, D., Persky, H., & Miller, J. (2008). The Nation’s Report Card: Writing 2007 (NCES 2008‐

468). Washington, DC: National Center for Educational Statistics. Institute for Educational

Sciences, US Department of Education.

Scarcella, R. (2003). Academic English: A conceptual framework. Los Angeles: University of California

Language Minority Research Institute.

Shoebottom (2008), General academic vocabulary [Web log post]. Retrieved from

http://esl.fis.edu/vocab/index-a.htm

Schwanenflugel, P. J., Stahl, S. A., & McFalls, E. L. (1997). Partial word knowledge and vocabulary

growth during reading comprehension. Journal of Literacy Research, 29(4), 531‐553.

Snodgrass, J. G., & Vanderwart, M. (1980). A standardized set of 260 pictures: Norms for name

agreement, familiarity and visual complexity. Journal of Experimental Psychology: Human

Learning & Memory, 6, 174-215.

Snow, C. E. (2010). Academic language and the challenge of reading for learning about science, Science,

328, 450 – 452.

http://esl.fis.edu/vocab/index-a.htm


Snow, C. E., & Uccelli, P. (2009). The challenge of academic language. In D.R. Olson & N. Torrance

(Eds.), Cambridge Handbook of Literacy. New York City: Cambridge University Press.

Snow, C., Lawrence, J., & White, C. (2009). Generating knowledge of academic language among urban

middle school students. Journal of Research on Educational Effectiveness, 2(4), 325‐344.

Stahl, S. A., & Fairbanks, M. M. (1986). The effects of vocabulary instruction: A model‐based meta‐

analysis. Review of Educational Research, 56, 72‐110.

Stahl, S. A., & Vancil, S. J. (1986). Discussion is what makes semantic maps work in vocabulary

instruction. The Reading Teacher, 40, 62‐69.

Stanovich, K.E. (1986). Matthew effects in reading: Some consequences of individual differences in the

acquisition of literacy. Reading Research Quarterly, 21, 360–407.

Townsend, D., & Collins, P. (2009). Academic vocabulary and middle school English learners: An

intervention study. Reading and Writing, 22, 993‐1020.

Valeri, L., & VanderWeele, T. J. (2013). Mediation analysis allowing for exposure–mediator interactions

and causal interpretation: Theoretical assumptions and implementation with SAS and SPSS

macros. Psychological Methods, 18(2), 137.

Vaughn, S., Martinez, L.R., Linan‐Thompson, S., Reutebuch, C.K., Carlson, C.D., & Francis, D.J. (2009).

Enhancing social studies vocabulary and comprehension for seventh‐grade English language

learners: Findings from two experimental studies. Journal of Research on Educational

Effectiveness, 2 (4), 297–324.

Wu, X., Anderson, R. C., Nguyen-Jahiel, K., & Miller, B. (2013). Enhancing motivation and engagement

through collaborative discussion. Journal of Educational Psychology, 105, 622-631.

Zhang, J., Anderson, R. C., & Nguyen-Jahiel, K. (2013). Language-rich discussions for English language

learners. International Journal of Educational Research, 58, 44-60.


Table 1

Means and standard deviations of use of academic vocabulary types in whale interview by intervention

condition

Condition

Variable

Collaborative

Groups

(N = 160)

Direct

Instruction

(N = 153)

Control

Condition

(N = 147)

Total occurrences of academic vocabulary 4.79 (3.51) 4.09 (3.13) 2.78 (2.33)

General academic vocabulary 3.14 (2.54) 2.57 (2.29) 1.93 (1.81)

Domain-specific

vocabulary

Ecosystem words 1.08 (1.12) 0.86 (1.00) 0.54 (0.79)

Economy words 0.04 (0.19) 0.08 (0.30) 0

Public policy words 0.54 (0.67) 0.58 (0.77) 0.31 (0.52)

Total occurrence 1.65 (1.51) 1.52 (1.50) 0.86 (1.05)


Table 2

Zero-order Pearson correlations between use of academic vocabulary and pretest language and social measures (N = 460)

1 2 3 4 5 6 7 8 9 10 11 12

1. Reading comprehension –

2. Object naming .36*** –

3. Talkativeness .11* .10* –

4. Good idea nominations .42*** .13** .32*** –

5. Leadership nominations .33*** .12** .11* .71*** –

6. Indegree centrality .16*** .04 .18*** .44*** .46*** –

7. Information centrality .19*** .13** .12* .22*** .22*** .53*** –

8. Betweenness centrality .02 .03 .11* .22*** .23*** .54*** .30*** –

9. Peer-liking rating .06 -.04 .12** .48*** .41*** .60*** .35*** .34*** –

10. General academic vocabulary .21*** .19*** .11* .09† .08† -.00 .12** .06 -.10* –

11. Domain-specific vocabulary .34*** .20*** .20*** .17*** .14** .08† .06 .03 -.02 .41*** –

12. Total academic vocabulary .31*** .23*** .17*** .14** .12** .03 .11* .06 -.08† .91*** .75*** –

Note. † p <.10, * p <.05, ** p < .01, *** p < .001


Table 3

Generalized Poisson mixed models of uses of general academic vocabulary in whale interview

Empty model Model 1 Model 2 Model 3 Model 4 Model 5 Model 6 Model 7

Fixed effects

Intercept 0.89***(0.06) 0.88***(0.06) 0.88*** (0.06) 0.86*** (0.06) 1.37***(0.17) 1.47***(0.18) 1.30***(0.19) 1.10***(0.18)

Language ability

Reading comprehension1

0.013***

(0.003)

0.009**

(0.003)

0.008*

(0.003)

0.009**

(0.003)

0.007*

(0.003)

0.008*

(0.004)

0.008*

(0.004)

Object naming2

0.008**

(0.002)

0.007**

(0.003)

0.007**

(0.003)

0.007**

(0.003)

0.007**

(0.003)

0.007**

(0.003)

Social characteristics

Talkativeness3 0.27** (0.10) 0.30**(0.10) 0.30**(0.10) 0.26* (0.10) 0.25** (0.10)

Peer-liking4 -0.17**(0.06) -0.23***(0.06) -0.30***(0.07) -0.32***(0.06)

Leadership5 0.41* (0.20) 0.39* (0.19) 0.43* (0.19)

Information centrality6 3.09* (1.23) 2.95** (1.00)

Intervention condition7

CG vs. DI 0.21* (0.09)

CG vs. Control 0.47***(0.10)

DI vs. Control 0.26** (0.10)

Random effects

Variance of Intercept 0.09** (0.03) 0.09** (0.03) 0.09** (0.03) 0.09** (0.03) 0.08** (0.03) 0.08** (0.03) 0.06** (0.02) 0.02† (0.01)

Fit Statistics

AIC 1981 1968 1961 1956 1948 1946 1942 1928

Note. 1 Class-mean-centered Gates-MacGinitie reading comprehension score (corrected for guessing). 2 Class-mean-centered object naming score (number of objects correctly named per minute). 3 Rate of classmates’ nominations for talkativeness minus rate of nominations for quietness. 4 Average of classmates’ ratings of ‘how much you like to play with’ this individual (five-point Likert scale) 5 Rate of classmates’ nominations for being a ‘good leader.’ 6 Weighted average of direct and indirect friendship ties. 7 CG refers to collaborative group work and DI refers to direct instruction. † p <.10, * p <.05, **p <.01, ***p <.001


Figure 1A. Path model of productive use of general academic vocabulary (GAV) in whale transfer

interview, excluding classroom dialogue.

* p <.05, ** p < .01, *** p < .001.

Figure 1B. Path model of productive use of general academic vocabulary (GAV) in whale transfer

interview, including classroom dialogue.

* p <.05, ** p < .01, *** p < .001.

Whale GAV

Control GAV

0.20***

c = 0.28* Collaborative Groups (=1) vs.

Direct Instruction (=0)

Wolf Print GAV

0.04***

Whale GAV Wolf Dialogue

GAV

Control GAV

0.18***

a = 0.66***

0.04***

Collaborative Groups (=1) vs

Direct Instruction (=0)

Wolf Print GAV

b = 0.15**

0.03***

c′ = 0.07


Appendix A. Whale Transfer Interview

Hunting whales is called whaling. People have been hunting whales for over a thousand years, but

now people have different opinions about whaling.

Some people say that we should hunt whales, because they are eating too many fish. Whales are

the largest animals living in the oceans, and they need to eat a huge amount of fish and other sea creatures

every day. For example, some whales may eat more than 8,000 pounds of food a day. Several kinds of

fish that whales eat are already disappearing, and only a few of them are left. Other people do not think

that whales are eating too much fish. These people argue that whales are not the only animals that eat fish.

Most of the fish are eaten by people, by other fish, and by seabirds. In fact, some kinds of whales do not

eat fish at all. Instead, they eat very small plants and other tiny animals.

Whales also have an effect on the businesses of different people. People who hunt whales argue

that whaling provides jobs to them and the people who work in restaurants. Whale hunters can make a lot

of money selling whales to restaurants. The meat from one whale can feed as many as 1,000 people for

almost two months. In some countries, like Norway, whale meat is a major source of food. Other people

say that whaling hurts whale-watching businesses. Each year, millions of people take tours to watch

whales in the ocean. These people spend money on boats, travel, hotels, and food. Whale-watching makes

a lot of money for people in many countries.

Some people are worried that whaling will affect whale populations. These people say that there

are not many whales left in the ocean, because too many whales have been killed by hunters and there is

less food for the whales to eat. They say that if we keep hunting whales, whales could disappear forever.

However, people who want whaling argue that not all kinds of whales are endangered. Some kinds of

whales have always been plentiful. Other kinds of whales were few in number in the past, but their

numbers are now increasing. Also, in countries like Norway and Japan, whaling is a tradition. Keeping

this tradition is very important for people in these countries.

Big Question:

Do you think we should allow people to hunt whales?

Prompt 1: Only if reasons are omitted

1. Tell me why you think we should [should not] allow people to hunt whales.

Prompt 2: Only if counter-argument is omitted

1. Could there be people who do not agree with you?

2. What would be their opinion and reasons?

Prompt 3: Only if rebuttal is omitted

1. What would you say to people who do not agree with your position?


Appendix B. Print exposure of domain-specific vocabulary in the Wolf Unit

Ecosystem

words (N = 25)

CG

ecosystem

textbook

DI

textbook

Economy

words (N = 17)

CG

economy

textbook

DI

textbook

Public policy

words (N = 18)

CG

public policy

textbook

DI

textbook

balance 12 7 agriculture 6 6 advocate 11 7

carnivore 9 7 compensation 5 6 biologist 12 8

competitor 3 4 compete 5 7 citizen 18 18

consumer 30 12 cost 5 5 common good 30 17

ecosystem 133 60 economy 68 30 community 23 20

endangered 6 11 expense 7 7 conflict 3 3

endurance 2 2 income 12 12 culture 2 2

extinct 5 4 livestock 41 35 heritage 2 2

food web 46 21 logger 20 7 interest(s) 22 26

global warming 4 5 lumber 6 5 majority 26 19

Habitat 3 8 manufacturing 4 4 minority 26 19

herbivore 6 4 permit (noun) 6 6 need(s) (noun) 1 1

nature 60 17 profit 4 4 opinion 36 20

natural resource 4 18 ranching 90 74 position 31 14

omnivore 6 4 service 6 6 public policy 33 14

population 33 26 timber 48 26 represent 17 11

predator 23 24 tourism 23 20 right(s) (noun) 17 15

prey 26 21 tradition 2 2

producer 26 12

recover 3 3

reintroduce 26 12

scavenger 6 7

species 41 6

starvation 1 2

territory 12 12


Appendix C. The first 50 most frequent general academic words in the Wolf Unit

reason increase passage except major

agree decrease source experience hardly

hurt lose deserve protect professional

affect survive worth destroy supply

instead since blame regular surface

business attack common action belong

disagree amount example hire choice

generate environment basic law continue

spend fault provide produce disturb

argument harm decision exist earn

Children’s Productive Use of Academic Vocabulary · Keywords: productive use of academic...

Documents

Transcript of Children’s Productive Use of Academic Vocabulary · Keywords: productive use of academic...